By Dave Benson, K1SWL

Taming the Trap Dipole

A
fter our recent move from a city article describes the development of a from the outer wires by a pair of traps ad-
location to several acres of self-supported 10/15/17 meter trap dipole. justed to 21.1 MHz. Although the dimen-
wooded bliss, it was only natural This project moved from the back burner sions shown are for the 10/15/17-meter
that a young man’s fancy would turn to to the “gotta try it” category when I found bands, there’s nothing to prevent you from
thoughts of . . . antennas! I’ve experi- that the local home-improvement emporium developing other combinations.
mented with any number of antenna con- carried 8-foot lengths of 3/8-inch aluminum The traps themselves are quite
figurations over the years, but multiband C-channel stock. This material has one im- simple—a parallel-resonant tuned circuit
operation always seemed to involve tun- portant advantage: all surfaces are flat, adjusted to the center of each amateur
ers used to press non-resonant wires into which eases a number of construction de- band of interest. I constructed each of
service. With the “clean slate” afforded tails. The joints between the element sec- these from iron-powder toroidal cores and
me with the new location, I decided I tions need to be an insulating material and a pair of silver mica capacitors. Each trap
wanted to pursue the “hook up the coax of sufficient strength to carry the weight of uses two 1 kV-rated capacitors in series
and forget it” approach. I’m also reluc- the outboard sections. The ideal material and T94 cores, the largest that would fit
tant to spend my limited discretionary for this application turned out to be 3/8-inch in the “low-profile” trap enclosures I
funds on commercial antennas when the square black Delrin (plastic) stock, which chose. I used Serpac C-series enclosures
homebrew approach works well. has good tensile strength properties.1 This available from mail-order distributors,
One approach to a multiband dipole material is also available in sizes up to 4 and a number of choices are also avail-
design is the so-called “fan dipole” inches square (at daunting prices) for ap- able through RadioShack. Figure 2 shows
wherein a separate electrical half-wave- plications where higher strength is required. the construction details—a pair of ma-
length of wire is added in parallel at the Figure 1 shows the dimensions of the chine screws exits through the rear wall
feedpoint for each band of interest. This trap antenna. The innermost dipole section of the trap enclosure and passes through
can become mechanically cumbersome (10 meters) is decoupled from the rest of holes drilled through the insulator stock
after the first several bands and interac- the antenna by a pair of traps tuned to 28.1 and the aluminum C-channel.
tion between bands becomes noticeable, MHz. The next pair of sections is decoupled The traditional tool for adjusting traps
at least with close wire spacings. I elected has been a grid-dip meter, and this has
instead to pursue the trap approach. This 1
Notes appear on page 30. been supplanted more recently by antenna

Figure 1—The dimensions of the trap antenna. Other dimensions can be devised for bands other than 10, 15 and 17 meters. At A,
the schematic of the trap. At B, dimensions for one side of the dipole antenna.
C1A, C1B—100 pF, 1 kV silver mica L1—10 meters: 9 turns on a T94-10 T94-6 toroidal core. The coils must be
capacitor. toroidal core; 15 meters: 11 turns on a tuned to resonance.
28 March 2002
 The Noise Bridge
Diode D1 is a source of broadband
noise. This noise is amplified to useful
levels by the two-stage circuit com-
prising Q1, Q2 and associated compo-
nents. Although there’s no attempt
made to frequency-compensate this
noise source, there’s plenty of signal
for our purposes—its output level
ranges from S9+20 dB at 1.8 MHz to
S7 at 30 MHz. In practice, when the
impedances connected to points B and
U are equal, this “bridge” circuit is in a
balanced condition and output to the
receiver is at a null. The only “tricky to its internal noise level alone. The
bit” in this circuit consists of the trifilar noise bridge is now adjusted for a
winding T1. [The circuit board project null—the impedance presented by the
offering uses color-coded wire for this 100-ohm resistance and stray capaci-
toroid, so hookup is pretty much fool- tance is now balanced by the bridge’s Figure B—How to hook up the noise
proof.] R1 and C1 settings. bridge.

So Now What? Putting it all Together cated the noise null. This null will be
Let’s put this to practical use: Con- If you add the trap—a parallel L-C fairly broad; however, it should be
nect a 100-Ω ¼ W resistor across the circuit—at its resonance frequency easy to locate using 1-MHz and then
“unknown” terminals and connect to across that 100-ohm resistor, there’d 100 kHz tuning steps.
your receiver with a length of coax. be no disturbance to the null since its Once you’ve found the null, bunch
Apply dc power (8-15 V) to the noise impedance at the intended operating the toroid turns together to lower the
bridge circuit and you should hear a frequency is theoretically infinite. trap resonance frequency or spread
loud rushing noise in the receiver. Away from the resonance frequency, the turns apart to raise the resonance
Adjust control R1 for minimum S- the noise level will rise as the receiver frequency. There’s a fair amount of
meter indication and then C1. Once is tuned off to either side. Finding the adjustment possible without resorting
these are both adjusted carefully, the trap’s resonant frequency amounts to to changing the toroid turns count—the
noise level in the receiver should drop tuning your receiver until you’ve lo- 21 MHz traps, for instance, could be
tuned in this manner to cover a range
of 19-22 MHz.
Caution
My initial attempts at repeatable
resonance measurements were incon-
sistent—the “casual” approach using
clip leads yielded well over a MHz of
variation in resonance frequency at 25
MHz! It’s critical to make the leads
from the “unknown” terminals on the
bridge to the traps as rigid as is practi-
cal. I used 2-inch lengths of no. 20
magnet wire to the 100-Ω parallel load
and installed solder lugs outboard of
that resistor. This allowed the traps to
be added and removed with a mini-
mum of change in stray capacitance,
which affects the resonance measure-
ment significantly. Once these precau-
tions were taken, the measurements
became reassuringly repeatable. Note:
Figure A—The schematic diagram of the noise bridge, based on a design that Once these trap hookup connections
appears in The ARRL Antenna Book. All resistors are 5%, ¼-W carbon composition. are ready to go and prior to adding the
D1—6.3-V, 0.5-W Zener diode, 1N753A or equiv. traps, be sure to readjust C1 for a
Q1, Q2—High-speed NPN switch, PN2222A, 2N4401 or equiv. noise null—this effectively tunes out
T1—4 turns trifilar-wound on FT37-43 toroid; observe phasing. the test setup stray capacitance.

analyzers. If you don’t have access to schematic diagram for this circuit, and this and working outward one band at a time.
either of these tools, though, despair not! is taken largely intact from The ARRL An- With a 4-inch spacing between the ends
If you have an HF transceiver with gen- tenna Book.2 A printed circuit-board kit was of the 8-foot channel sections, the 10-
eral coverage capability, you’ve already developed as a club project and is available meter antenna simply worked on the first
got most of what you need. to interested builders.3 try. Resonance for this dipole was at 28.1
The remaining piece of equipment re- MHz and SWR characteristics were fairly
quired is a noise bridge. Despite the arcane- Antenna Adjustment broad due to the element thickness.
sounding name, this is a simple circuit that This antenna was developed by start- Upon adjustment of a pair of 10-meter
is easily duplicated. The sidebar shows the ing with the innermost (10-meter) section traps, these were added to the element
March 2002 29
 output or less, that’s not everyone’s “cup
of tea.” I’ve tested this antenna at 100 W
without incident. EZNEC analysis using
the published “Q” values for the toroid
trap material shows antenna gain at 28
MHz to be 0.8 dB down from the expected
free-space values, and 0.9 dB down at 21
MHz. At 18.1 MHz, the loss is approxi-
mately 0.25 dB. These values would be
somewhat improved with the use of
Figure 2—Construction details of the trap. See text. higher-Q inductors. This design has
traded “compact” and “low-profile” for
modest gain penalties—proof indeed of
rary rod sections with C-channel and
the old adage about “no free lunch.”
added a pair of 15-meter traps. With the
A point of interest—I calculate the peak
addition and adjustment of the outer 17-
voltage across the traps at that power level
meter sections, this completed the design
to be over 1 kV. This is no place for junkbox
for my applications, so I elected to leave
capacitors of questionable pedigree! A
the outer antenna ends in the form of ¼-
high-quality NP0 capacitor type is a
inch rod stock to reduce weight and lower
“must”—the types typically available from
the antenna’s visible “profile.”
your local electronics emporium may be
The center insulator/mounting block is quite lossy at high frequencies, and this will
constructed from a 3/8 × 3 × 12-inch block translate into considerable component heat-
of Delrin plastic. This provides sufficient ri- ing and disappointing performance. The
gidity for this antenna, although if the con- 500-V silver mica capacitors available from
Figure 3—The SWR characteristics for cept is extended to lower bands you’d
the trap dipole. Since the author operates the large distributors are sufficient for
probably want thicker plastic material. A lower-power (QRP) operation.6
primarily CW and data modes, the
lengths are optimized for the lower end of small plastic box at the feedpoint contains a I installed this antenna at the 35-foot
each band. choke balun. I constructed this using a short level above my roof and have been very
length of RG-174 coax looped three times pleased with its performance. After years
ends and outboard sections for 15 meters
through a group of six FT37-43 ferrite tor- of “low-profile” QRP operation, my suc-
were added. Rather than use the C-chan-
oids. There’s nothing magical about this ap- cess rate snagging contacts on the first
nel material on the initial adjustments, I
proach—any of a number of other methods call has improved markedly. To a large
found it much more convenient to install
can be used to achieve the same goal. extent, the old maxim of “Put it up high
outboard sections of ¼ inch aluminum rod
stock. This material proved to be quite Construction and in the clear” applies here! As a final
easy to trim to length with a pair of bolt- “food-for-thought” consideration, the
All fastening hardware for the trap di- trap-construction scheme I’ve described
cutters! Tune-up was done at an initial
pole should be of stainless steel, and toothed would lend itself nicely to multiband ver-
height of 20 feet. Element lengths are
lock washers are needed to maintain integ- tical and ground-plane antennas.
adjusted using an SWR bridge and trans-
mitter to determine the frequency at rity of the tightened joints. Once the traps
which SWR is minimum and adjusting are adjusted to the desired resonance fre- Acknowledgments
accordingly. A gentle suggestion: It’s quencies, the trap enclosures are sealed shut Special thanks to Seabury Lyon,
much easier to start “long” and subtract with an edge-bead of model airplane cement AA1MY, for his assistance with the noise
material rather the reverse! and resonance was re-checked. This final bridge project.
You’ll find that the outboard lengths check ensures that adding the enclosure Notes
for each additional lower-frequency band covers has not disturbed the trap frequen- 1
Delrin plastic may be purchased in small
do not meet the familiar formula for cies—a possibility given the tight quarters quantities from McMaster-Carr, www.
afforded by the enclosures I chose. mcmaster.com; see “raw materials.”
computing dipole lengths. The traps 2
The ARRL Antenna Book, 19 th Ed., p 27-24.
themselves present a very high impedance Results 3
A noise bridge kit consisting of double-sided/
at their design frequency but below this silkscreened printed-circuit board, on-board
frequency are inductive. This has the ef- The SWR characteristics for this antenna parts and RG-174/U cable with BNC con-
are shown in Figure 3. I operate primarily nector and instructions is available from
fect of shortening the resonant length of the New England QRP Club for $17 ($20
the antenna. [It has a modest effect in low- CW and data modes, so my interest is in the overseas) postpaid. Checks or money
ering feedpoint resistance as well, but is lower end of each band; the lengths in this orders payable to S. Lyon, AA1MY, 99
not significant within the context of this article reflect that preference. Whatever fre- Sparrowhawk Mtn Rd, Bethel, ME 04217.
4
application.] With the trap components I quency you choose, you know you’ve done Dean Straw, N6BV, Ed., The ARRL Antenna
a careful job tuning the traps if the addition Book, 19th Ed., p 6-28.
chose, each outboard section length was 5
EZNEC is available from Roy Lewallen,
shortened by 30-35% over the expected of these traps and outboard sections has no W7EL, www.eznec.com.
values for a dipole. For the adventurous, effect on resonance frequency of the inner 6
Toroids are available from Amidon Associ-
this length may be estimated by calculat- antenna portion. Their presence, though, will ates (tel 714-850-4660) or Palomar Engi-
narrow the effective SWR bandwidth as you neers, www.palomar-engineeers.com.
ing the effective impedance of the trap at 1-kV silver mica capacitors are available
the lower band and applying it to any of move away from resonance—the trap-an- from RF Parts Co (www.rfparts.com; tel
several tools. This information is found tenna bandwidths are lower than for that of 800-737-2787).
in graphical form in The ARRL Antenna a “plain-vanilla” dipole.
Book4 or by use of EZNEC.5 Dave Benson, K1SWL, is a frequent contribu-
With the length of the 15-meter sec- Trap Losses tor to QST. He can be reached at dave@
tion under control, I replaced the tempo- Although I normally operate at 5 W smallwonderlabs.com.
30 March 2002